Sains Malaysiana 49(11)(2020): 2609-2623 http://dx.doi.org/10.17576/jsm-2020-4911-01

Intraspecific Phenotypic Variation in Nearly Threatened Mottled , (Hamilton, 1822) (Variasi Fenotip Intrakhusus Patung Belang Nandus yang Hampir Terancam, Nandus nandus (Hamilton, 1822))

MD. SAROWER-E-MAHFUJ, MD. ABDUS SAMAD, FEE FAYSAL AHMED, MD. ABDUL ALIM, YOSNI BAKAR & SIMON KUMAR DAS*

ABSTRACT Understanding intraspecific phenotypic plasticity is a prerequisite of stock identification, evolutionary studies, sustainable utilization, and fishery conservation. In this study, intraspecific phenotypic plasticity was assessed in terms of the external features (i.e. meristic, morphometric, and truss-based morphometrics) of the wild Nandus populations from four freshwater sources in Southwestern Bangladesh. Fish samples were collected from Arial Kha River, Madaripur (AKRM, n=26); Nabaganga River, Jhenaidah (NRJ, n=22); Bohnni Baor, Gopalganj (BBG, n=26); and Dhakuria Beel, Jashore (DBJ, n=22). Meristic, morphometric, and truss network data were subjected to one-way ANOVA followed by post hoc (Tukey-HSD) test. The meristic counts of all the samples demonstrated significant differences only in one of the six characters. By contrast, significant differences were observed in 8 morphometric characters and 31 truss network data from 16 morphometric characters and 35 truss network data, respectively. Principal component (PCA) and canonical variate analyses (CVA) were also performed on morphometric and truss-based network data. Meristic and morphometric results from PCA and CVA showed that populations were completely intermingled, forming a compact cluster within intrapopulation levels, while truss morphometric characters formed a separate cluster. Three dendrograms independently based on phenotypic relationships among the individuals of the four populations also confirmed the absence of phenotypic differentiation among the population due to clustering of different groups. The baseline information resulting from the current study would be useful for genetic studies and further in situ conservation of Nandus populations in Bangladesh. Keywords: Canonical variate analysis; freshwater; morphometric; meristic; nandus; principle component analysis; Truss morphometry

ABSTRAK Memahami keplastikan fenotip intrakhusus adalah satu pra-syarat untuk mengenal pasti stok, kajian evolusi dan pemanfaatan lestari dan pemuliharaan dalam perikanan. Dalam kajian ini kepelbagaian fenotip intrakhusus dinilai dari segi ciri luaran (iaitu meristik, morfometri dan morfometri dasarkan truss) daripada populasi liar ikan nandus dari empat sumber air tawar di selatan-barat Bangladesh. Sampel ikan dikumpulkan dari Arial Kha River, Madaripur (AKRM), (n = 26); Sungai Nabaganga, Jhenaidah (NRJ), (n = 22); Bohnni Baor, Gopalganj (BBJ), (n = 26); dan Dhakuria Beel, Jashore (DBJ), (n = 22). Data meristik, morfometri dan rangkaian truss dianalisis menggunakan varians satu arah (ANOVA) diikuti dengan ujian Post-hoc (Tukey-HSD). Perhitungan meristik untuk kesemua sampel menunjukkan perbezaan yang signifikan hanya dalam satu ciri daripada enam ciri manakala perbezaan yang signifikan diperhatikan dalam 8 ciri morfometrik dan 31 rangkaian data truss masing-masing daripada 16ciri morfometrik dan 35 rangkaian data truss. Di samping itu, analisis komponen utama (PCA) dan analisis fungsi diskriminasi (CVA) dilakukan dengan menggunakan morfometrik dan data rangkaian berasaskan truss. Hasil daripada PCA dan CVA menunjukkan populasi terpisah sepenuhnya serta membentuk kelompok yang padat dalam tahap intra- INTRODUCTION populasi. Tiga dendrogram secara bebas berdasarkan hubungan fenotip antara individu daripada empat populasi River dibina. Populasi NRJ, BBG dan DBJ membentuk populasi kumpulan masing-masing berdasarkan meristik, morfometrik dan truss morfometrik. Maklumat asas yang dihasilkan daripada kajian semasa adalah mudah untuk kajian genetik dan pemuliharaan populasi Nandus secara in situ di Bangladesh. Kata kunci: Air tawar; analisis fungsi diskriminasi; meristik; morfometrik; morfometri Truss; nandus 2610

INTRODUCTION Nandus is a freshwater fish commonly known as mud perch or mottled nandus and considered a small Phenotypic plasticity is the ability of an organism to adjust indigenous in Bangladesh (Ross et al. 2003). This its body maintenance in response to genetic-environmental fish species is widely distributed in fresh and brackish interactions. Sometimes, phenotypic plasticity, phenotypic waters, including ditches, ponds, beels (saucer-shaped responsiveness, flexibility, and condition sensitivity are perennial water bodies), and inundated fields throughout entirely synonymous in evolutionary biology (West- South Asian countries (Ahmed 2008; Rahman 2005). Eberhard 1989). The plethora of outcomes, such as changes Nandus is a carnivorous organism that entirely feeds on in body shape and size, allometry, feeding habits, sexual larvae and insects, crustaceans, filamentous algae, and dimorphism, and behavioral and physiological states, small fishes (Agarwal & Sharma 1966). Although this can be collectively or solely achieved from phenotypic species is considered a bony fish that survives at a low plasticity after a certain period of time (Langerhans oxygen level, it can camouflage when any prey, small fish, 2008). Thus, similar to other organisms with this and even a predator is present in a water body (Mustafa property, fishes are not an exception. Fishes also exhibit et al. 1980). This fish also plays a substantial role in the an outstanding extent of variation in their external body overall nutrition for poor-rural-living and low-income- shape morphologies, such as meristic and morphometric generating communities in Bangladesh (Das & Zamal characters, at a species level (Oufiero & Whitlow 2016). 2000). According to IUCN-Bangladesh (Chowdhury Consequently, morphometrics can be defined as an array of 2015), this species is categorized as nearly threatened quantitative analyses, such as biological outline, or shape because of habitat destruction, overexploitation, disparity among organisms with respect to environmental anthropogenic activities, and climate change (Rahman factors (Webster & Sheets 2010). Moreover, studies on 2005). As such, morphometric and meristic studies should the morphogenesis of fishes plays a fundamental role in be conducted to detect intraspecific phenotypic plasticity evolutionary analysis and proper management (Başusta et and ensure sustainability in the future. al. 2014; Kalhoro et al. 2015). At present, no adequate information regarding the Information related to the stock structure analysis of intraspecific phenotypic variation in N. nandus in the a species or a population is a prerequisite of the expansion freshwaters of Bangladesh is available. Therefore, this of proper biodiversity management and conservation study aimed to investigate the intraspecific phenotypic (Turan et al. 2005). Morphological dissimilarities are variations in N. nandus based on meristic, morphometric, observable characteristics in a fish or a fish population and truss network system. and caused by genetic factors, genetic-environmental interactions, and abiotic and biotic influences (Crispo MATERIALS AND METHODS 2008; Silva et al. 2013). Generally, in early developmental stages, fishes express their phenotypic plasticity in two FISH SAMPLING ways, that is, isometric size variation due to growth and allometric shape variation caused by developmental A total of 100 individuals of Nandus sp. were collected alteration (Cadrin 2000). Freshwater fishes exhibit a high from four different freshwater sources in Bangladesh degree of body shape variation because of physiological from September 2017 to November 2017: Arial Kha and environmental conditions, resulting in genetic River, Madaripur (AKRM); Nabaganga River, Jhenaidah variation and phenotypic plasticity (Eklöv & Svanbäck (NRJ); Bohnni Baor, Gopalganj (BBG); and Dhakuria 2005). Numerous techniques, such as morphometrics Beel, Jashore (DBJ) (Figure 1 & Table 1). The samples and meristics, traditional tags, otolith microchemistry, were placed in an ice box and immediately brought into and electronic tags, have been extensively used for stock the Laboratory of Fish Biology and Aquaculture, Jashore identification. Morphometric traits are one of the most University of Science and Technology, Bangladesh. The used and cost-effective methods to detect intraspecific minimum and maximum total lengths (TL) of the fish phenotypic variation in species (Mir et al. 2013). Naturally, specimens were 6.94 and 12.89 cm, respectively. fishes undergo ontogeny in an allometric pattern from the beginning of their life cycle (Hood & Heins 2000; Svanbäck COUNTING OF MERISTIC CHARACTERS & Eklöv 2002). To reinforce the inherent limitation of In six meristic characters, the numbers of dorsal spiny conventional morphometric approaches, the truss-networks fin rays (DSFR), dorsal soft fin rays (SFR), caudal fin rays formed by two or more interconnecting distances across- (CFR), anal fin rays (AFR), pelvic fin rays (PevFR), and body that ultimately produced chronological sequence pectoral fin rays(P ecFR) were counted in each sample by of associated polygons has been progressively utilized using magnifying glasses and needles. (Strauss & Bookstein 1982). 2611

FIGURE 1. Map of Bangladesh showing collection sites of N. nandus from four freshwater sources

TABLE 1. Sampling details of N. nandus from four freshwater sources in Bangladesh

Serial no. Populations Abbreviations Locations Number of Mean SL in specimens cm (SD) 1 Arial Kha River, Madaripur AKRM 23.23°N 90.18 °E 26 9.55 (0.54)

2 Nabaganga River, Jhenaidah NRJ 23.54°N 89.17 °E 22 7.76 (0.91)

3 Bohnni Baor, Gopalganj BBG 23.16◦N 89.21 °E 26 7.38 (1.26)

4 Dhakuria Beel, Jashore DBJ 23.16◦N 89.21 °E 26 8.42 (1.19)

MEASUREMENT OF MORPHOMETRIC AND TRUSS background. Then, the individual fish was categorized NETWORKS with a definite code for documentation. A Cybershot First, the image of the samples was digitized after the DSC-W730 digital camera (Sony, China) was used to fish were thawed under running tap water, wiped well, capture digital images, which provided a whole record of and placed on a smooth platform with a white paper as a body shape and allowed re-measurements when necessary 2612

(Cadrin & Friedland 1999). The morphometrics and truss created on each fish image, which was constructed by distances from the digital images of the specimens were interconnecting 35 truss network measurements (Figure extracted using tpsDig2V2.1 (Rohlf 2006; Table 2). In 2). the case of truss network distances, 13 landmarks were

TABLE 2. Seventeen morphometric characters were used for the analysis intra/specific phenotypes of mottledN. nandus

Characters Description Total length (TL) Distance from the tip of the lower jaw to the longest caudal fin ray Standard length (SL) Distance from the tip of the lower jaw to the end of the vertebral column Pre-dorsal length (PDL) Front of the lower lip to the origin of the first ray of the first dorsal fin Post orbital head length (POL) Distance from the posterior margin of the eye to the end of the operculum Pre-pectoral length (PPCL) Front of the lower lip to the origin of the pectoral fin Pre-pelvic length (PPVL) Front of the lower lip to the origin of thwwe pelvic fin Length of the first dorsal fin base (LDFB1) From base of first dorsal fin ray to base of last dorsal fin ray Length of the second dorsal fin base(LDFB 2) From base of the second dorsal fin ray to base of last dorsal fin ray Length of anal fin base(LAFB) From base of the first anal fin ray to base of the last anal fin ray Upper jaw length (UJL) Straight line measurement between the snout tip and posterior edge of maxilla Straight line measurement between the snout tip and posterior edge of Lower jaw length (LJL) mandible Body depth (BD) Maximum depth measured from the base of the first dorsal fin ray Snout length (SNL) The front of the upper lip to the fleshy anterior edge of the orbit Eye diameter (ED) The greatest crystal-like diameter of the orbit Distance between front of the lower lip to the posterior end of the opercular Head length (HD) membrane Depth of caudal peduncle (DCP) The least depth of the tail base Inter orbital (IO) Distance between dorsal side of both eyes

FIGURE 2. Location of 13 anatomic landmarks of N. nandus for constructing 35 truss networks on fish body illustrated as close circle (black). The descriptions of landmarks are follows: (1) anterior tip of the upper snout, (2) forehead (end of the frontal bone), (3) origin of the first dorsal fin, (4) endpoint of the first dorsal fin, (5) origin of the second dorsal fin, (6) endpoint of the second dorsal fin, (7) dorsal origin of caudal fin, (8) ventral origin of the caudal fin, (9) endpoint of the anal fin, (10) origin of anal fin,1) (1 endpoint of the pelvic fin, (12) down of the operculum, and (13) anterior tip of the lower snout 2613

DATA ANALYSES and DBJ populations were similar and NRJ population All original morphometric and truss data were subjected significantly differed from BBG and DBJ populations, to general descriptive analysis to check their normality while AKRM population was intermediate. The differences before they were further examined using SPSS version 21 (P > 0.05) in DSFR (F = 1.558, P > 0.05), SFR (F = 2.335, (SPSS, Chicago, IL, USA). An allometric formula, which P > 0.05), CFR (F = 0.765, P > 0.05), AFR (F = 1.058, P was described by Elliott et al. (1995) and slightly > 0.05), and PevFR (F = 1.058, P > 0.05) among the four modified in the present study, was used to remove the size populations were not statistically significant (Table 3). effect from the dataset based on (1): Eight morphometric characters (i.e., SL, PDL, PPVL, LDFB1, LAFB, UJL, BD, and HL) also significantly varied b (P < 0.05) among 16 morphometric characters (Table Madj = M (Ls / Lo) (1) 4). For instance, SL (F= 2.898, P < 0.05) of the AKRM and DBJ populations were highly significant to each where M is the original measurement; Madj is the size- other, whereas the BBG and NRJ populations were adjusted measurement; Lo is the TL of the fish; sL is the overall mean of the TL of all the fish from all the samples; intermediate among the four populations. In case of PDL and b is estimated as the slope of the regression of logM (F = 3.870, P < 0.05), the AKRM and BBG populations resembled similar and showed significant difference on logLo by using all the fish samples in all the populations for each character from the observed data. Meristic, from DBJ population, conversely NRJ population was morphometric, and truss distance data were compared intermediate among the four populations. Similarly, for among populations via one-way ANOVA followed by PPVL (F = 6.740, P < 0.05), DBJ population showed post hoc (Tukey-HSD) test. Size-adjusted data were significant disparity compared to the three remaining also subjected to principal component analysis (PCA) populations of AKRM, BBG, and NRJ. Additionally, and discriminant function analysis (canonical variate LDFB1 (F = 3.700, P < 0.05) character showed significant analyses (CVA)). All statistically analyzed data were disparity between BBG and DBJ populations, whilst considered using a probability of P=0.05. Three separate AKRM and NRJ populations exhibited intermediate among dendrograms with a complete linkage and a Euclidean the four populations. Moreover, LAFB (F = 5.868, P < distance were drawn using meristic, morphometric, and 0.05) character of AKRM population possessed significant truss morphometric data. The entire statistical analyses difference from BBG and DBJ populations while NRJ were performed using SPSS version 21 (SPSS, Chicago, population exhibited as intermediate. The UJL (F = 6.220, IL, USA) and R version 3.5.2. P < 0.05) character of BBG and NRJ populations showed significant differences to each other, but AKRM and DBJ populations remained intermediate and equally similar to RESULTS each other. Furthermore, the BD (F = 4.116, P < 0.05) and Mean values were compared through one-way ANOVA HL (F = 20.299, P < 0.05) characters showed significant followed by Tukey-HSD post hoc test of each meristic, differences in AKRM, BBG, and DBJ populations to each morphometric, and truss morphometric character from four other even though the NRJ population showed intermediate. wild Nandus populations (Tables 3 to 5, respectively). In meristic characters, PecFR (F = 7.182, P < 0.05) of the BBG

TABLE 3. Comparison of the (mean ± SD) of meristic characters of N. nandus in four populations namely, Arial Kha river, Madaripur (AKRM); Bohnni baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria beel, Jashore (DBJ) in Bangladesh

Meristic characters AKRM BBG NRJ DBJ F P-value DSFR 12.15 ± 1.12 12.44 ± 0.72 12.73 ± 0.72 12.48 ± 0.96 1.558 0.205

SFR 11.53 ± 1.10 10.72 ± 1.31 11.41 ± 1.00 11.40 ± 1.35 2.355 0.077

CFR 13.23 ± 0.71 13.28 ± 0.89 13.36 ± 1.04 13.60 ± 1.08 0.768 0.515

AFR 10.15 ± 1.43 9.68 ± 0.90 9.73 ± 1.42 10.16 ± 1.25 1.058 0.371

PevFR 6.62 ± 1.03 6.28 ± 0.89 7.23 ± 1.99 6.64 ± 0.95 2.230 0.090

PecFR 12.31 ± 1.15ab 11.36 ± 2.03b 13.13 ± 1.08a 11.48 ± 1.44b 7.182 0.000* *P < 0.05 . SD: Standard deviation. F: The ratio of between-group variability and within group variability in one-way analysis of variance (ANOVA). Different small superscripts in each row differs the values of meristic characters 2614

TABLE 4. Comparison of the (mean ± SD) of morphometric characters of N. nandus in four populations namely, Arial Kha river, Madaripur (AKRM); Bohnni baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria beel, Jashore (DBJ) in Bangladesh

Morphometric characters AKRM BBG NRJ DBJ F P-value SL 8.19 ± 0.37b 8.31 ± 0.41ab 8.23 ± 0.23ab 8.44 ± 0.27a 2.898 0.039* PDL 4.03 ± 0.27a 3.77 ± 0.21a 3.88 ± 0.37ab 3.68 ± 0.61b 3.870 0.012* POL 2.21 ± 0.63 2.45 ± 0.71 2.26 ±0.57 2.56 ± 0.61 1.686 0.175 PPCL 3.32 ± 0.33 3.20 ± 0.25 3.34 ± 0.29 3.33 ± 0.34 1.178 0.322 PPVL 2.86 ± 0.28b 2.71 ± 0.48b 2.75 ± 0. 39b 3.18 ± 0.48a 6.740 0.000* LDFB1 3.08 ± 0.34ab 2.88 ± 0.30b 3.07 ± 0.32ab 3.25 ± 0.57a 3.700 0.014* LDFB2 0.72 ± 0.13 0.75 ± 0.16 0.77 ± 0.20 0.82 ± 0.17 0.166 0.919 LAFB 1.14 ± 0.11a 0.97 ± 0.18b 1.04 ± 0.22ab 0.94 ± 0.20b 5.868 0.001* UJL 0.88 ± 0.23bc 0.82 ± 0.16c 1.15 ± 0.51a 1.13 ± 0.37ab 6.220 0.001* LJL 0.85 ± 0.23 0.96 ± 0.26 1.01 ± 0.40 1.29 ± 0.56 1.546 0.208 BD 3.02 ± 0.15a 2.69 ± 0.21b 2.78 ± 0.42ab 2.64 ± 0.67b 4.116 0.009* SNL 0.73 ± 0.16 0.64 ± 0.13 0.82 ± 0.49 0.68 ± 0.08 1.803 0.152 ED 0.69 ± 0.10 0.68 ± 0.24 0.66 ± 0.11 0.72 ± 0.13 0.688 0.588 HL 2.45 ± 0.45b 1.77 ± 0.73c 2.94 ± 0.85ab 3.13 ± 0.65a 20.299 0.000* DCP 0.99 ± 0.09 1.06 ± 0.33 1.03 ± 0.08 1.08 ± 0.10 0.966 0.412 IO 1.16 ± 0.04 1.17 ± 0.03 1.19 ± 0.18 1.20 ± 0.09 0.671 0.572 * P < 0.05. SD: Standard deviation. F: The ratio of between-group variability and within group variability in one-way analysis of variance (ANOVA). Different small superscripts in each row differs the values of morphometric characters

In truss morphometric characters, out of 35 4.413, P < 0.05) character proved significant differences morphometric characters 31 showed significant differences in BBG and DBJ populations than the NRJ population (Table 5). The characters 2-3 (F = 38.546, P < 0.05), 4-5 whereas AKRM population showed intermediate among (F = 18.408, P < 0.05), 7-8 (F = 20.082, P < 0.05), 8-9 (F the three remaining populations. Likewise, 2-10 (F = = 12.050, P < 0.05), 9-10 (F = 20.139, P < 0.05), 11-12 6.829, P < 0.05) character showed significant difference (F = 16.641, P < 0.05), 1-11 (F = 8.416, P < 0.05), 2-12 in AKRM and DBJ populations whilst BBG and NRJ (F = 7.675, P < 0.05), 3-12 (F = 28.377, P < 0.05), 3-11 populations showed intermediate among the three (F = 14.315, P < 0.05), 3-10 (F = 13.878, P < 0.05), 4-11 remaining populations. Correspondingly, 3-9 (F = 18.693, (F = 7.415, P < 0.05), 6-9 (F = 3.614, P < 0.05), 2-9 (F = P < 0.05) character demonstrated significant difference in 11.030, P < 0.05), and 1-9 (F = 31.212, P < 0.05) of the DBJ DBJ population than the NRJ and BBG populations but population demonstrated highly significant differences the AKRM population exhibited intermediate between from those of the three remaining populations. In addition, BBG and NRJ populations. Together with, 4-10 (F = 10-11 (F = 8.567, P < 0.05) and 1-3 (F = 9.874, P < 0.05) 7.107, P < 0.05) character proved significant difference characters of DBJ population significantly differed from in BBG population than NRJ population while AKRM and the three remaining populations. Similarly, 5-6 (F = DBJ populations remained intermediate between BBG 13.271, P < 0.05) character showed significant difference and NRJ populations. Additionally, 6-11 (F = 4.641, P < in NRJ population from the three remaining populations 0.05) character demonstrated significant difference in of AKRM, BBG, and DBJ. AKRM and DBJ populations but BBG and NRJ populations On the flip of site, 3-4 (F = 9.915, P < 0.05) character showed intermediate among the four populations. Equally, demonstrated significant differences in BBG and DBJ 1-10 (F = 53.819, P < 0.05) character showed significant populations whereas AKRM and NRJ populations remained differences in DBJ, AKRM and BBG populations whilst intermediate among the four populations. Similarly, 6-7 (F NRJ population remained intermediate state among = 5.046, P < 0.05) character showed significant difference the populations. Furthermore, 7-11 (F = 13.271, P < in NRJ population than the BBG and DBJ populations 0.05) character of DBJ population showed significant while AKRM population remained intermediate among deviation than the BBG population while AKRM and NRJ the three remaining populations. Additionally, 2-11 (F = populations showed intermediate among the populations. 2615

TABLE 5. Comparison of the (mean ± SD) of truss morphometric characters of N. nandus in four populations namely, Arial Kha river, Madaripur (AKRM); Bohnni baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria beel, Jashore (DBJ) in Bangladesh

Characters AKRM BBG NRJ DBJ F P-value 1-2 2.98 ± 0.33 3.08 ± 0.65 2.96 ± 0.71 3.26 ± 0.41 1.572 0.201 2-3 1.00 ± 0.38b 0.87 ± 0.19b 0.99 ± 0.17b 2.32 ± 1.00a 38.546 0.000* 3-4 2.56 ± 0.42bc 2.20 ± 1.00c 2.87 ± 0.54ab 3.34 ± 0.95a 9.915 0.000* 4-5 0.51 ± 0.21b 0.50 ± 0.17b 0.56 ± 0.46b 1.97 ± 1.50a 18.408 0.000* 5-6 0.59 ± 0.16b 0.55 ± 0.11b 0.78 ± 0.28a 0.45 ± 0.14b 13.271 0.000* 6-7 0.88 ± 0.08ab 0.79 ± 0.13b 0.93 ± 0.14a 0.79 ± 0.21b 5.046 0.003* 7-8 0.95 ± 0.06b 1.01 ± 0.14b 1.01 ± 0.14b 1.23 ± 0.17a 20.082 0.000* 8-9 0.93 ± 0.24b 0.87 ± 0.09b 1.05 ± 0.21b 1.26 ± 0.37a 12.050 0.000* 9-10 1.29 ± 0.57b 1.04 ± 0.36b 1.03 ± 0.31b 2.18 ± 0.94a 20.139 0.000* 10-11 2.18 ± 0.51a 2.38 ± 0.54a 2.23 ± 0.59a 1.69 ± 0.39b 8.567 0.000* 11-12 1.71 ± 0.51b 1.54 ± 0.22b 1.63 ± 0.38b 2.38 ± 0.67a 16.641 0.000* 12-1 2.13 ± 0.32 2.09 ± 0.18 1.96 ± 0.27 1.95 ± 0.62 1.364 0.264 12-13 2.46 ± 0.88a 2.18 ± 0.65ab 1.67 ± 0.74b 2.79 ± 1.06a 7.288 0.000* 1-3 3.84 ± 0.62a 3.84 ±0.33a 3.90 ± 0.61a 3.14 ± 0.71b 9.874 0.000* 1-11 3.61 ± 0.69b 3.34 ± 0.42b 3.36 ± 0.51b 4.08 ± 0.70a 8.416 0.000* 2-12 2.89 ± 0.22b 2.98 ± 0.57b 2.83 ± 0.45b 3.42 ± 0.58a 7.675 0.000* 2-11 3.07 ± 0.45ab 3.33 ± 0.79a 2.87 ± 0.41b 3.41 ± 0.54a 4.413 0.006* 2-10 4.19 ± 0.63a 3.75 ± 0.30bc 4.18 ± 0.60ab 3.59 ± 0.70c 6.829 0.000* 3-12 3.36 ± 0.35b 3.08 ± 0.24b 3.23 ± 0.34b 4.03 ± 0.57a 28.377 0.000* 3-11 3.00 ± 0.23b 3.10 ± 0.57b 3.12 ± 0.63b 3.81 ± 0.50a 14.315 0.000* 3-10 3.80 ± 0.18b 3.60 ± 0.33b 3.65 ± 0.65b 4.27 ± 0.41a 13.878 0.000*

3-9 3.92 ± 0.24bc 3.56 ± 0.33c 4.14 ± 0.47b 4.60 ± 0.81a 18.693 0.000* 4-11 3.78 ± 0.59b 3.58 ± 0.48b 3.55 ± 0.64b 4.34 ± 0.91a 7.415 0.000* 4-10 2.56 ± 0.41bc 3.28 ± 0.89a 2.54 ± 0.68c 3.07 ± 0.71ab 7.107 0.000* 4-9 2.39 ± 0.69b 1.60 ± 0.29a 2.04 ± 0.47b 2.37 ± 0.53b 13.511 0.000* 6-9 1.40 ± 0.37b 1.39 ± 0.17b 1.50 ± 0.36b 1.64 ± 0.32a 3.614 0.016* 6-8 1.39 ± 0.11 1.46 ± 0.18 1.46 ± 0.18 2.10 ± 0.54 1.388 0.251

7-9 1.47 ± 0.13b 2.24 ± 1.40a 1.71 ± 0.71ab 1.83 ± 0.78ab 3.410 0.021*

6-11 3.99 ± 0.33a 3.41 ± 0.63ab 3.62 ± 1.15ab 2.96 ± 1.51b 4.641 0.004*

6-10 2.23 ± 0.27 2.38 ± 1.15 2.30 ± 0.79 2.13 ± 1.16 0.371 0.774 2-9 4.83 ± 0.89b 4.76 ± 0.92b 4.35 ± 0.93b 5.95 ± 1.30a 11.030 0.000* 1-4 6.36 ± 0.78b 5.11 ± 0.72c 6.01 ± 0.91b 7.83 ± 1.27a 36.578 0.000* 1-10 6.13 ± 0.34b 5.31 ± 0.67c 5.86 ± 1.00ab 7.98 ± 1.03a 53.819 0.000* 1-9 6.99 ± 0.47b 5.92 ± 0.82c 6.78 ± 0.78b 8.22 ± 1.21a 31.212 0.000* 7-11 4.82 ± 0.35ab 4.61 ± 0.57b 4.74 ± 0.52ab 5.14 ± 0.72a 4.233 0.007* * P< 0.05. SD: Standard deviation. F: The ratio of between-group variability and within group variability in one-way analysis of variance (ANOVA). Different small superscripts in each row differs the values of truss morphometric characters 2616

Multivariate analyses (i.e. PCA and CVA) were were 0.526, 0.577, and 0.810 for meristic, morphometric, performed using meristic, morphometric, and truss and truss morphometric characters, respectively, and morphometric data to detect the exact causes of variation Bartlett’s test of sphericity showed significant results (P in the specimens of the four populations. However, the < 0.05). According to Kaiser (1974), these KMO values insufficient sample size is a major bottleneck of the fish can be ranked as moderate (0.5-0.7), good (0.7-0.8), and morphology studies during multivariate analysis. In this excellent (0.8-0.9). Therefore, the obtained results from case, a ratio of sample size (N) among all specimens KMO and Bartlett’s tests suggested that the extracted data and the number of characters (F) of at least 2.8-3.5 was from each sample were highly fit for the factor analysis considered (Kocovsky et al. 2009; Parsons et al. 2003). of meristic, morphometric, and truss morphometric Insignificant N values may fail to adequately capture co- characters. variance or morphological variation, possibly leading to In the PCA of six meristic characters, three factors false conclusions regarding changes among populations with eigenvalues higher than 1 were extracted, and the (McGarigal et al. 2000). However, in the present study, the remaining factors were discarded. The results elucidated total number of specimens was 100 (N), and the numbers 62.79% of the total variance. The first, second, and third of meristic, morphometric, and truss morphometric principal components (PC1, PC2, and PC3, respectively) characters were 6 (P), 16 (P), and 35 (P), respectively. described 25.8, 19.9, and 17.1% of the variance, Through the use of N and P values, the ultimate ratios respectively (Table 6). Among the three PCs, the most were 16.66 (N:P) for meristic parameters, 6.25 (N:P) significant loadings on PC1 were AFR, DSFR, SFR, CFR, for morphometric parameters, and 2.85 (N:P) for truss and PecFR (Table 6). CVA produced three canonical morphometric parameters, respectively. Consequently, variations (CV; i.e., CV1, CV2, and CV3) for six meristic PCA and CVA were performed to examine the characters characters. CV1, CV2, and CV3 accounted for 72.2, (meristic, morphometric, and truss morphometrics) that 18.6, and 9.2% of group variability, respectively (Table mostly discriminated the populations. Before conducting 6). Pooled within-group correlations between canonical the final PCA, data were validated with Bartlett’s test variables and CVs showed the following contributions of of sphericity, and the Kaiser–Meyer–Olkin (KMO) the six characters: PecFR to CV1, DSFR and SFR to CV2, measurement was performed. The statistical range of the and CFR and PevFR to CV3 (Table 6). KMO values varied between 0 and 1. The KMO values

TABLE 6. Component loadings of first three principal components(PC) and canonical covariates (CV) for meristic characters in N. nandus collected from Arial Kha river, Madaripur (AKRM); Bohnni baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria beel, Jashore (DBJ) in Bangladesh. Character descriptions are given in material and methods section

PCA CVA

Meristic characters PC 1 PC 2 PC 3 CV 1 CV 2 CV 3

PecFR 0.415 0.561 0.439 0.742* -0.037 -0.189

DSFR 0.603 0.361 -0.043 0.116 -0.583* 0.390

SFR 0.527 -0.613 -0.261 0.271 0.535* 0.524

AFR 0.686 -0.347 -0.023 -0.024 0.500* 0.363

CFR 0.468 0.383 -0.362 -0.049 -0.061 0.663*

PevFR 0.209 -0.327 0.795 0.382 -0.142 0.408*

Eigenvalue 1.546 1.195 1.026 0.412 0.106 0.530

Variance % 25.8 19.9 17.1 72.2 18.6 9.2

Cumulative % 25.8 45.7 62.8 72.2 90.8 100.0

* Largest absolute correlation between each variable and any canonical variate function 2617

In the PCA of 16 morphometric characters, three and CV3) for 16 morphometric characters; that is, CV1, factors with eigenvalues higher than 2 were extracted, CV2, and CV3 accounted for 64.9, 25.5, and 9.6% of group and the remaining factors were discarded. These results variability, respectively (Table 7). Pooled within-group elucidated 40.54% of the variance. PC1, PC2, and PC3 correlations between canonical variables and CVs showed accounted for 17.7, 13.4, and 9.44% of the distinction, the following contributions among 16 morphometric respectively. Among the three PCs, the most significant characters: HL, LJL, and LDFB2 to CV1; LAFB, BD, PDL, loadings on PC1 were HL, BD, PPCL, PPVL, UJL, LDFB1, SNL, SL, POL, PPCL, and DCP to CV2; and PPVL, UJL, and ED (Table 7). CVA produced three CVs (CV1, CV2, LDFB1, ED, and IO to CV3 (Table 7).

TABLE 7. Component loadings of first three principal components(PC) and canonical covariates (CV) for morphometric characters in N. nandus collected from Arial Kha river, Madaripur (AKRM); Bohnni baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria beel, Jashore (DBJ) in Bangladesh. Character descriptions are given Table 2

PCA CVA

Characters PC 1 PC 2 PC 3 CV 1 CV 2 CV 3

HL 0.598 0.010 -0.391 0.598* 0.501 0.009

LJL 0.225 0.464 -0.515 0.173* -0.113 -0.006

LDFB2 0.059 0.110 0.005 0.055* -0.041 -0.022

LAFB 0.337 -0.644 -0.078 0.188 0.069* 0.267

BD 0.579 -0.321 0.205 -0.161 0.360* 0.326

PDL 0.149 -0.758 0.136 -0.180 0.351* 0.207

SNL 0.017 -0.079 -0.268 0.025 0.287* -0.227

SL 0.326 0.361 0.511 0.188 -0.271* 0.085

POL 0.146 0.380 0.544 0.109 -0.257* 0.026

PPCL 0.724 -0.389 0.025 0.101 0.202* 0.040

DCP 0.164 0.288 -0.070 0.085 -0.186* -0.083

PPVL 0.668 0.293 0.151 0.329 -0.053 0.533*

UJL 0.458 0.380 -0.443 0.331 0.191 -0.332*

LDFB1 0.654 0.026 -0.097 0.251 0.147 0.281*

ED 0.423 0.143 0.333 0.073 -0.073 0.234*

IO 0.060 0.242 0.235 0.112 -0.009 -0.129*

Eigenvalue 2.829 2.144 1.513 1.389 0.544 0.205

Variance % 17.681 13.401 9.458 64.9 25.5 9.6

Cumulative % 17.681 31.082 40.540 64.9 90.4 100.0

* Largest absolute correlation between each variable and any discriminant function 2618

In the PCA of 35 truss morphometric characters, three morphometric characters. CV1, CV2, and CV3 accounted factors with eigenvalues greater than 2 were extracted, for 58.4, 27.6, and 14.0% of group variability (Table and the remaining factors were discarded. The results 8). Pooled within-group correlations between canonical elucidated 56.80% of the variance. PC1, PC2, and e4q3 variables and CVs showed the following contributions described 35.8, 12.20, and 8.80% of the distinction, among 35 truss morphometric characters: 22 characters respectively (Table 8). The most noteworthy loadings on (1-10, 2-3, 1-4, 1-9, 3-12, 9-10, 4-5, 11-12, 3-10, 3-9, 2-9, PC1 were 1-2, 2-3, 1-4, 1-9, 3-12, 9-10, 4-5, 11-12, 3-10, 1-3, 1-11, 8-9, 10-11, 4-11, 2-12, 12-13, 7-11, 6-9, 6-8 and 3-9, 2-9, 1-11, 8-9, 4-11, 2-12, 12-13, 7-11, 6-9, 3-4, 7-8, 6-10) to CV1; 2 characters (5-6 and 3-4) to CV2; and 11 4-9, 4-10, 3-1, 2-11, and 1-2 (Table 8). CVA yielded three characters (7-8, 4-9, 4-10, 2-10, 3-11, 6-11, 7-9, 6-7, 2-11, canonical variations (CV1, CV2, and CV3) in 35 truss 1-2 and 12-1) to CV3 (Table 8).

TABLE 8. Component loadings of first three principal components (PC) and canonical covariates (CV) for truss morphometric characters in N. nandus collected from Arial Kha river, Madaripur (AKRM); Bohnni baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria beel, Jashore (DBJ) in Bangladesh. Character descriptions are given in material and methods section

PCA CVA Characters PC 1 PC 2 PC 3 CV 1 CV 2 CV 3 1-10 0.829 -0.299 0.115 -0.464* 0.140 0.078 2-3 0.709 -0.425 -0.233 -0.383* 0.080 0.229 1-4 0.750 -0.384 0.248 -0.374* 0.165 -0.066 1-9 0.753 -0.394 0.319 -0.340* 0.177 -0.059 3-12 0.860 0.028 0.143 -0.339* 0.081 0.059 9-10 0.819 -0.137 -0.184 -0.288* 0.005 0.087 4-5 0.680 -0.331 -0.343 -0.258* 0.046 0.200 11-12 0.791 -0.127 -0.077 -0.258* 0.047 0.100 3-10 0.733 0.247 0.045 -0.240* 0.027 0.035 3-9 0.786 -0.184 0.359 -0.234* 0.223 0.046 2-9 0.731 0.197 -0.056 -0.205* -0.064 0.100 1-3 -0.461 0.574 0.243 0.192* 0.001 -0.136 1-11 0.690 -0.031 0.284 -0.188* 0.005 0.001 8-9 0.521 -0.032 0.088 -0.186* 0.157 0.133 10-11 -0.336 0.538 0.195 0.184* -0.061 -0.039 4-11 0.776 0.088 0.204 -0.176* -0.006 0.036 2-12 0.558 0.386 -0.141 -0.161* -0.034 0.156 12-13 0.542 0.130 -0.050 -0.151* -0.124 -0.044 7-11 0.719 0.364 0.203 -0.130* 0.043 -0.001 6-9 0.635 0.007 0.004 -0.099* 0.081 0.096 6-8 0.212 -0.082 -0.129 -0.068* 0.011 0.068 6-10 0.101 0.544 -0.286 0.037* -0.012 0.022 5-6 -0.237 0.300 0.289 0.169 0.231* -0.086 3-4 0.649 0.066 0.221 -0.164 0.173* 0.047 7-8 0.802 0.160 -.158 -0.242 0.052 0.321* 4-9 0.483 -0.282 0.354 -0.173 0.112 -0.297* 4-10 0.494 0.396 -0.206 -0.027 -0.160 0.267* 2-10 0.029 0.349 0.777 0.093 0.100 -0.254* 3-11 0.701 0.368 -0.263 -0.215 0.045 0.232* 6-11 -0.138 0.155 0.803 0.082 -0.001 -0.231* 2619

7-9 0.239 0.771 -0.372 0.030 -0.078 0.210* 6-7 0.126 0.399 0.499 0.059 0.153 -0.166* 2-11 0.579 0.553 -0.303 -0.077 -0.129 0.141* 1-2 0.488 0.556 -0.093 -0.066 -0.027 0.088* 12-1 -0.099 0.416 0.155 0.032 -0.080 -0.083* Eigenvalue 12.531 4.259 3.088 7.439 3.516 1.780 Variance % 35.8% 12.2% 8.8% 58.4 27.6 14.0 Cumulative % 35.8% 47.9% 56.8% 58.4 86.0 100.0 *Largest absolute correlation between each variable and any discriminant function

The biplot arrangements, that is, PC1 versus PC2 four populations in CV1 versus CV2 (Figure 3(f)). Three and CV1 versus CV2, of the meristic (Figure 3(a) and dendrograms were constructed on the basis of the complete 3(d)), morphometric (Figure 3(b) and 3(e)), and truss linkage and Euclidean distance to examine the phenotypic morphometric (Figure 3(c) and 3(f)) characters were relationships independently among the individuals of the constructed using PCA and CVA results, respectively. The four populations. In the dendrogram, intermingling results biplot results of the meristic characters demonstrated four were observed in the individuals in meristic characters, and multivariate spaces with a significant overlap and unclear the individuals of the NRJ population mainly contributed differentiation among the four populations (Figure 3(a) and as the distinct population (Figure 4(a)). Similarly, 3(d)). The biplot results of the morphometric characters individuals were also performed as intermixing stage by exhibited four multivariate spaces with a high overlap using morphometric characters, where BBG population among the four populations in PC1 versus PC2 (Figure mainly formed as distinct population (Figure 4(b)). 3(b)) and a slight overlap in the result of CV1 versus CV2 Consequently, distinct outcomes were also demonstrated (Figure 3(e)). The biplot results of the truss morphometric by the individuals in truss morphometric characters, and characters displayed four multivariate spaces with a the DBJ population diverged as a unique distinct population slight overlap in PC1 versus PC2 (Figure 3(c)), whereas (Figure 4(c)). distinct separation was observed in individuals from the

EA B C

D E F

FIGURE 3. (a-c) Principal component analysis, and (d-f) and canonical variate analysis of Nandus nandus obtained from meristic, morphometric, and truss morphometric characters, respectively. Fish samples collected from Arial Kha river, Madaripur (AKRM); Bohnni baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria beel, Jashore (DBJ) in Bangladesh 2620

A B C

FIGURE 4. Dendrogram with complete linkage and Euclidean distance of meristic, morphometric and truss morphometric data of Nandus nandus: (a) dendrogram derived from meristic data, (b) dendrogram derived from morphometric data, and (c) dendrogram derived from truss morphometric data. Fish samples collected from Arial Kha river, Madaripur (AKRM); Bohnni Baor, Gopalganj (BBG); Nabaganga river, Jhenidah (NRJ) and Dhakuria Beel, Jashore (DBJ) in Bangladesh

DISCUSSION AND CONCLUSION could be assigned to conjoined genetic bases and ecological Among all vertebrates fishes are one of the most variations that originated in topographical juxtaposition susceptible organisms that pose high environmentally (Saborido-Rey & Nedreaas 2000; Walsh et al. 2001). induced morphological dissimilarities. Hence, fishes Nevertheless, high deviations in PecFR may have been exhibit maximum phenotypic plasticity among populations caused by the effect of environmental influences of other organisms, even though the same species occupy formed at the time of ontogenetic development through a single ecological niche (Allendorf 1987; Wimberger pre- or post-fecundation influence (Lindsey 1988). The 1992). However, our study disclosed the intraspecific discrepancy of PecFR may be ascribed to the nature of the phenotypic plasticity of Nandus in a large range from number of fin rays, which are static in later stages than four freshwater ecological sources of Southwestern other meristic characters over ontogeny (Akbarzadeh et Bangladesh. Similarly, Goswami and Dasgupta (2007) al. 2009). The difference in the number of rays of pectoral studied meristic characters and observed that the average fins may be due to the temperature in their ecological numbers of fin rays are in the range of 12-13 for DSFR, niches and feeding modes (Kahilainen & Østbye 2006; 16 for PecFR, 15 for CFR, and 7-9 for AFR. Significant Trabelsi 2002). Conversely, the consequences of individual results have also been observed in N. oxyrhynchus from polymorphism and quantitative genetics on meristic the Mekong Basin in Vietnam (Ng et al. 1996), N. prolixus variations are not omitted. from Northeastern Borneo in Indonesia (Chakrabarty et al. In the present study, the differences in morphometric 2006), and N. meni from the Noakhali Coast in Bangladesh and truss measurements were highly significant in post (Hossain & Sarker 2013). The meristic characters used hoc tests among the four populations. Such a degree of in this research (i.e. DSFR, CFR, AFR, PevFR, and PecFR) phenotypic changes among the populations may be due 2621 to their distinct geographical site, current environmental results in Macrognathus pancalus, Xenentodon cancila, dissimilarity of the four ecological niches, or different and Lepidocephalichthys guntea, respectively. The descendants. Generally, fishes and aquatic organisms exhibit divergent of wild group based on morphological data has high sensitivity to environmental changes and rapidly alter been hypothesized to be formed due to environmental and their body shapes with respect to their new environmental genetic factors (Allendorf & Phelps 1988; Nakamura et conditions for proper adaptation. Phenotypic characters al. 2003; Okomoda et al. 2018; Solomon et al. 2015). The can exhibit high plasticity because of the fluctuation of finding of this study may just attest to similarity of origin environmental conditions, such as several abiotic (e.g. of the different wild populations understudied. temperature, water quality parameters, and climate change) The finding of this study are highly useful as a basis and biotic (e.g. food abundance, host–pathogen–parasite for conducting further studies on Nandus populations. interaction) factors (Allendorf & Phelps 1988; Solomon For aquaculture and open-water fishery management, et al. 2015; Wimberger 1992;). Usually, fishes are highly the information obtained in this study may be helpful vulnerable because of environment-induced morphological in sorting out superior populations after further studies variations in comparison with other vertebrate within are performed. More so, further studies, such as genetic intra- and interpopulation levels (Allendorf et al. 1987; research and analysis on the influences of environmental Wimberger 1992). However, describing the cause of dynamics, are required for the in situ and ex-situ the morphological changes between/among populations conservation and artificial seed propagation of certain (Cadrin 2000) is difficult when certain observed variances populaces to protect and save this nearly threatened species are due to growth differences, mortality, and reproduction from extinction. rates (Silva et al. 2013). The phenotypic plasticity of fish is high because they adapt their physiological ACKNOWLEDGEMENTS characteristics and behavior to environmental changes, We would like to thank anonymous reviewers for and such adaptations eventually alter their morphological their very useful comments that greatly improved the traits (Stearns 1983). Morphological alterations in aquatic earlier version of this manuscript. Thanks to UKM for vertebrates with minimal environmental differences may the financial support through the research grant ‘GP- be difficult to distinguish by studying gross morphometric 2019-K019059’ and UKM-Sime Darby Foundation and meristic characters only. Therefore, truss network Chair in Climate Change Grant ‘ZF-2019-003’ to the dimensions were included in this trial. Turan et al. (2004) corresponding author. indicated that truss network systems are dominant tools in fish stock identification and stock delineation. In the present research, the truss network system might be REFERENCES efficiently used to differentiate the four populations. Ahmed, M.S. 2008. Assessment of fishing practices on the Highly significant variations were anticipated because exploitation of the Titas floodplain in Brahmanbaria, of four entirely different ecological niches (i.e. the two Bangladesh. Turkish Journal of Fisheries and Aquatic rivers are open water habitats, and the two beels are Sciences 8(2): 329-334. closed water habitats). Ecologically or environmentally Agarwal, V.P. & Sharma, K.U. 1966. Studies on the physiology of digestion in Nandus nandus (Ham.). Proceedings of the persuaded phenotypic discrepancies may be beneficial to Indian Academy of Sciences 64: 157-168. the investigation of the stock structure of exploited species, Akbarzadeh, A., Facxsarahmand, H., Shabani, A.A., Karami, M., particularly during a short time frame (Gain et al. 2017; Kaboli, M., Abbasi, K. & Rafiee, G.R. 2009. Morphological Hossain et al. 2010; Mahfuj et al. 2017; Simon et al. 2010). variation of the pikeperch Sander lucioperca (L.) in the Phenotypic differentiation in the four populations southern Caspian Sea, using a truss system. Journal of showed strong overlap according to PCA and CVA Applied Ichthyology 25(5): 576-582. results. Morphometric and truss morphometric characters Allendorf, F.W. & Phelps, S.R. 1988. Loss of genetic variation usually play a significant role in the creation of stock in a hatchery stock of cutthroat trout. Transactions of the discrimination rather as compared to meristic characters. American Fisheries Society 109(5): 537-543. However, in this study, the four wild population could not Allendorf, F.W., Ryman, N. & Utter, F. 1987. Genetics and be separated into individually distinct multivariate spaces fishery management: Past, present and future in population genetics and fisheries management. In Population Genetics judging from the observation of overlap in PCA and the and Fishery Management, edited by Ryman, N. & Utter, F. dendrogram with complete linkage. This is contrary to the Washington: University of Washington Press. pp. 1-20. findings of Okomoda et al. (2018a, 2018b) who reported Başusta, A., Özer, E.I., Girgin, H., Serdar, O. & Başusta, that the pure and reciprocal crosses Clarias gariepinus N. 2014. Length-weight relationship and condition and Pangasianodon hypophthalmus can be discriminated factor of Hippocampus hippocampus and Hippocampus using morphological data. This also not in tandem with guttulatus inhabiting Eastern Black Sea. Pakistan Journal the finding of Hossain et al. (2010) with Labeo calbasu. of Zoology 46(2): 447-450. Mahfuj et al. (2019a, 2019b, 2019c) detected similar 2622

Cadrin, S.X. 2000. Advances in morphometric identification of lizardfish, Saurida tumbil from Pakistani waters. Pakistan fishery stocks. Reviews in Fish Biology and Fisheries 10(1): Journal of Zoology 47(4): 921-935. 91-112. Kocovsky, P.M., Adams, J.V. & Bronte, C.R., 2009. The effect of Cadrin, S.X. & Friedland, K.D. 1999. The utility of image sample size on the stability of principal component analysis processing techniques for morphometric analysis and stock of truss-based fish morphometrics.Transaction of American identification. Fisheries Research 43(1): 129-139. Fisheries Society 138: 487-496. Chakrabarty, P., Oldfield, R.G. & Ng, H.H. 2006. Nandus Langerhans, R.B. 2008. Predictability of phenotypic differentiation prolixus, a new species of leaf fish from northeastern Borneo across flow regimes in fishes. Integrative and Comparative (Teleostei: Perciformes: ). Zootaxa 1328: 51-61. Biology 48(6): 750-768. Chowdhury, G.W. 2015. Nandus nandus. The IUCN Red List of Lindsey, C.C. 1988. 3 factors controlling meristic variation. Fish Bangladesh. Volume 5: Freshwater Fishes. Dhaka: IUCN, Physiology. Vol. 11. pp. 197-274. Massachusetts: Academic International Union for Conservation of Nature, Bangladesh Press. Country Office. Mahfuj, M.S., Khatun, A., Boidya, P. & Samad, M. 2019a. Crispo, E. 2008. Modifying effects of phenotypic plasticity on Meristic and morphometric variations of barred spiny eel interactions among natural selection, adaptation and gene Macrognathus pancalus populations from Bangladeshi flow.Journal of Evolutionary Biology 21(6): 1460-1469. freshwaters: An insight into landmark-based truss network Das, M. & Zamal, N. 2000. Domestication of an endangered system. Croatian Journal of Fisheries: Ribarstvo 77(1): 7-18. fish speciesNandus nandus (Ham.) pt. 1. Laboratory rearing Mahfuj, M.S., Rahman, M.M., Islam, M., Samad, M.A., of young fish up to sexual maturity. Bangladesh Journal of Paul, A.K. & Adhikary, R.K. 2019b. Landmark-based Fisheries Research 4(2): 135-140. morphometric and meristic variations of freshwater garfish, Eklöv, P. & Svanbäck, R. 2005. Predation risk influences adaptive Xenentodon cancila from four natural stocks of South- morphological variation in fish populations. The American Western Bangladesh. Journal of Advanced Veterinary and Naturalist 167(3): 440-452. Research 6(1): 117-124. Elliott, N.G., Haskard, K. & Koslow, J.A. 1995. Morphometric Mahfuj, S.E., Rahman, S.U. & Samad, A. 2019c. Landmark- analysis of orange roughy (Hoplostethus atlanticus) off the based truss morphometrics delineate the stock structure continental slope of southern Australia. Journal of Fisheries of Lepidocephalichthys guntea. Journal of Fisheries and Biology 46(1): 202-220. Aquatic Science 14(1): 25-32. Gain, D., Mahfuj, M.S., Huq, K.A., Islam, S.S., Minar, M.H., Mahfuj, S., Ashraful, M.A., Parvez, I., Minar, M.H. & Samad, A. Goutham-Bharathi, M.P. & Das, S.K. 2017. Landmark-based 2017. Morphological variations of Labeo bata populations morphometric and meristic variations of endangered mrigal (Teleostei: Cyprinidae) in six rivers of Bangladesh: A carp, Cirrhinus cirrhosus (Bloch 1795), from wild and landmark-morphometric contribution. Iranian Journal of hatchery stocks. Sains Malaysiana 46(5): 695-702. Ichthyology 4(3): 270-280. Goswami, S. & Dasgupta, M. 2007. Analysis of the morphometric McGarigal, K., Cushman, S. & Stafford, S. 2000. Multivariate and meristic characters of the fish Nandus nandus Statistics for Wildlife and Ecology Research. New York: (Hamilton) from the new alluvial zone of West Bengal. Springer Verlag. Records of the Zoological Survey of India 107: 81-90. Mir, J.I., Sarkar, U.K., Dwivedi, A.K., Gusain, O.P. & Jena, J.K. Hamilton, F. 1822. An Account of the Fishes Found in the 2013. Stock structure analysis of Labeo rohita (Hamilton, River Ganges and Its Branches. Uttarakhand: Bishen Singh 1822) across the Ganga basin (India) using a truss network Mahendra Pal Singh. system. Journal of Applied Ichthyology 29(5): 1097-1103. Hood, C.S. & Heins, D.C. 2000. Ontogeny and allometry Mustafa, G., Ahmed, A.T.A. & Islam, K.R. 1980. Food, feeding of body shape in the blacktail shiner, Cyprinella habits and fecundity of a freshwater perch, meni fish. venusta. Copeia 2000(1): 270-275. Bangladesh Journal of Agriculture 5(4): 205-210. Hossain, M.A., Nahiduzzaman, M., Saha, D., Khanam, M.U.H. Nakamura, T. 2003. Meristic and morphometric variations in & Alam, M.S. 2010. Landmark-based morphometric and fluvial Japanese charr between river systems and among meristic variations of the endangered carp, Kalibaus Labeo tributaries of a river system. Environmental Biology of Fishes calbasu, from stocks of two isolated rivers, the Jamuna and 66(2): 133-141. Halda, and a hatchery. Zoological Studies 49(4): 556-563. Ng, H.H., Vidthayanon, C. & Ng, P. 1996. , Hossain, M.S. & Sarker, S. 2013. New species of leaf fishNandus a new species of leaf fish (Teleostei: Nandidae) from the meni (Perciformes: Nandidae) from Noakhali, Bangladesh. Mekong Basin. The Raffles Bulletin of Zoology44(1): 11-19. Zoology and Ecology 23(3): 191-197. Okomoda, T.V., Koh, I.C.C., Hassan, A., Amornsakun, T. & Kahilainen, K. & Østbye, K. 2006. Morphological differentiation Shahreza, S.M. 2018a. Morphological characterization of the and resource polymorphism in three sympatric whitefish progenies of pure and reciprocal crosses of Pangasianodon Coregonus lavaretus (L.) forms in a subarctic lake. Journal hypophthalmus (Sauvage, 1878) and Clarias gariepinus of Fish Biology 68(1): 63-79. (Burchell, 1822). Scientific Reports 8(1): 1-13. Kaiser, H.F. 1974. An index of factorial simplicity. Psychometrika Okomoda, V.T., Koh, I.C.C., Hassan, A., Amornsakun, T. & 39: 31-36. Shahreza, M.S. 2018b. Performance and characteristics Kalhoro, M.A., Liu, Q., Valinassab, T., Waryani, B., Abbasi, of the progenies from the crosses of Pangasianodon A.R. & Memon, K.H. 2015. Population dynamics of greater 2623

hypophthalmus (Sauvage, 1878) and Clarias gariepinus Turan, C., Ergüden, D., Gürek, M. & Turan, F. 2004. Genetic and (Burchell, 1822). Aquaculture 489(3): 96-104. morphologic structure of Liza abu (Heckel, 1843) populations Oufiero, C.E. & Whitlow, K.R. 2016. The evolution of phenotypic from the rivers Orontes, Euphrates and Tigris. Turkish Journal plasticity in fish swimming.Current Zoology 62(5): 475-488. of Veterinary and Animal Sciences 28(4): 729-734. Parsons, K.J., McWhinnie, K., Pilakouta, N. & Walker, L. 2020. Walsh, M.G., Bain, M.B., Squiers, T., Waldman, J.R. & Wirgin, I. Does phenotypic plasticity initiate developmental bias? 2001. Morphological and genetic variation among shortnose Evolution & Development 22(1-2): 56-70. sturgeon Acipenser brevirostrum from adjacent and distant Rahman, A.K.A. 2005. Freshwater Fishes of Bangladesh. 2nd ed. rivers. Estuaries 24(1): 41-48. pp. 394. Dhaka: University of Dhaka. Webster, M. & Sheets, H.D. 2010. A practical introduction to Rohlf, F.J. 2006. TPS Software Series. New York: Department landmark-based geometric morphometrics. Quantitative of Ecology and Evolution, State University of New York, Methods in Paleobiology 16: 168-188. Stony Brook. West-Eberhard, M.J. 1989. Phenotypic plasticity and the origins Ross, N., Islam, M.M. & Thilsted, S.H. 2003. Small indigenous of diversity. Annual Review of Ecology and Systematics fish species in Bangladesh: Contribution to vitamin A, 20(1): 249-278. calcium and iron intakes. The Journal of Nutrition 133(11): Wimberger, P.H. 1992. Plasticity of fish body shape. The effects 4021S-4026S. of diet, development, family and age in two species of Saborido-Rey, F. & Nedreaas, K.H. 2000. Geographic variation Geophagus (Pisces: Cichlidae). Biological Journal of the of Sebastes mentella in the Northeast Arctic derived from a Linnean Society 45(3): 197-218. morphometric approach. ICES Journal of Marine Science 57(4): 965-975. Md. Sarower-E-Mahfuj, Md. Abdus Samad & Md. Abdul Alim Simon, K.D., Bakar, Y., Temple, S.E. & Mazlan, A.G. 2010. Department of Fisheries and Marine Biosciences Morphometric and meristic variation in two congeneric Jashore University of Science and Technology archer fishesToxotes chatareus (Hamilton 1822) and Toxotes Jashore-7408 jaculatrix (Pallas 1767) inhabiting Malaysian coastal waters. Bangladesh Journal of Zhejiang University Science B 11(11): 871-879. Silva, S.E., Silva, I.C., Madeira, C., Sallema, R., Paulo, O.S. Fee Faysal Ahmed & Paula, J. 2013. Genetic and morphological variation in Department of Mathematics two littorinid gastropods: Evidence for recent population Jashore University of Science and Technology expansions along the East African coast. Biological Journal Jashore-7408 of the Linnean Society 108(3): 494-508. Bangladesh Solomon, S.G., Okomoda, V.T. & Ogbenyikwu, A.I. 2015. Intraspecific morphological variation between cultured and Yosni Bakar & Simon Kumar Das* wild Clarias gariepinus (Burchell) (Clariidae, Siluriformes). Department of Earth Sciences and Environment Archives of Polish Fisheries 23(1): 53-61. Faculty of Science and Technology Stearns, S.C. 1983. A natural experiment in life-history evolution: Universiti Kebangsaan Malaysia field data on the introduction of mosquito fishGambusia ( 43600 UKM Bangi, Selangor Darul Ehsan affinis) to Hawaii. Evolution 37(3): 601-617. Malaysia Strauss, R.E. & Bookstein, F.L. 1982. The truss: Body form reconstructions in morphometrics. Systematic Biology 31(2): Simon Kumar Das* 113-135. Marine Ecosystem Research Center Svanbäck, R. & Eklöv, P. 2002. Effects of habitat and food Faculty of Science and Technology resources on morphology and ontogenetic growth trajectories Universiti Kebangsaan Malaysia in perch. Oecologia 131(1): 61-70. 43600 UKM Bangi, Selangor Darul Ehsan Trabelsi, M. 2002. Le complexe Atherina boyeri. Caractérisation Malaysia biométrique, biochimique et génétique. Mise en de deux nouvelles espèces d’athérines dans le méditerranéen. *Corresponding author; email: [email protected] Université de Tunis 291. Turan, C., Yalçin, S., Turan, F., Okur, E. & Akyurt, I. 2005. Received: 12 February 2018 Morphometric comparisons of African catfish, Clarias Accepted: 22 May 2020 gariepinus, populations in Turkey. Folia Zoologica 54(1/2): 165-172.